U.S. patent application number 12/791526 was filed with the patent office on 2010-12-02 for listening device providing enhanced localization cues, its use and a method.
This patent application is currently assigned to Oticon A/S. Invention is credited to Marcus Holmberg, Thomas Kaulberg, Michael Syskind PEDERSEN.
Application Number | 20100303267 12/791526 |
Document ID | / |
Family ID | 41280397 |
Filed Date | 2010-12-02 |
United States Patent
Application |
20100303267 |
Kind Code |
A1 |
PEDERSEN; Michael Syskind ;
et al. |
December 2, 2010 |
LISTENING DEVICE PROVIDING ENHANCED LOCALIZATION CUES, ITS USE AND
A METHOD
Abstract
The invention relates to a listening device comprising an
ear-part adapted for being worn in or at an ear of a user, a front
and rear direction being defined relative to a person wearing the
ear-part in an operational position. The invention further relates
to a method of operating a hearing instrument, to its use, to a
listening system, to a computer readable medium and to a data
processing system. The object of the present invention is to
provide localization cues for indicating a direction of origin of a
sound source. The problem is solved in that the listening system
comprises (a) a microphone system comprising at least two
microphones each converting an input sound to an electrical
microphone signal, (b) a DIR-unit comprising a directionality
system for providing a weighted sum of the at least two electrical
microphone signals thereby providing at least two directional
microphone signals having maximum sensitivity in spatially
different directions and a combined microphone signal, and (c) a
frequency shaping-unit for modifying the combined microphone signal
to indicate directional cues of input sounds originating from at
least one of said spatially different directions and providing an
improved directional output signal. This has the advantage of
providing an alternative or an addition to natural localization
cues. The invention may e.g. be used in listening devices, e.g.
hearing instruments, head phones, headsets or active ear plugs.
Inventors: |
PEDERSEN; Michael Syskind;
(Smorum, DK) ; Holmberg; Marcus; (Smorum, DK)
; Kaulberg; Thomas; (Smorum, DK) |
Correspondence
Address: |
BIRCH STEWART KOLASCH & BIRCH
PO BOX 747
FALLS CHURCH
VA
22040-0747
US
|
Assignee: |
Oticon A/S
Smorum
DK
|
Family ID: |
41280397 |
Appl. No.: |
12/791526 |
Filed: |
June 1, 2010 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
61183483 |
Jun 2, 2009 |
|
|
|
Current U.S.
Class: |
381/313 |
Current CPC
Class: |
H04R 25/407
20130101 |
Class at
Publication: |
381/313 |
International
Class: |
H04R 25/00 20060101
H04R025/00 |
Foreign Application Data
Date |
Code |
Application Number |
Jun 2, 2009 |
EP |
09161700.0 |
Jun 2, 2009 |
EP |
09161700.1 |
Claims
1. A listening device comprising an ear-part adapted for being worn
in or at an ear of a user, a front and rear direction being defined
relative to a person wearing the ear-part in an operational
position, the listening device further comprising (a) a microphone
system comprising at least two microphones each converting an input
sound to an electrical microphone signal, (b) a DIR-unit comprising
a directionality system for providing a weighted sum of the at
least two electrical microphone signals thereby providing at least
two directional microphone signals having maximum sensitivity in
spatially different directions and a combined microphone signal,
and (c) a frequency shaping-unit for modifying the combined
microphone signal to indicate directional cues of input sounds
originating from at least one of said spatially different
directions and providing an improved directional output signal.
2. A listening device according to claim 1 comprising an analogue
to digital converter unit providing said electrical microphone
signals as digitized electrical microphone signals.
3. A listening device according to claim 1 comprising a
TF-conversion unit for providing a time-frequency representation of
the at least two microphone signals, each signal representation
comprising corresponding complex or real values of the signal in
question in a particular time and frequency range.
4. A listening device according to claim 3 wherein the DIR-unit is
adapted to detect from which of said spatially different directions
a particular time frequency range originates.
5. A listening device according to claim 1 wherein the frequency
shaping unit is adapted to move the directional cues of a given
time frequency range to a relatively lower frequency range.
6. A listening device according to claim 5 where differences in the
directional microphone signals attributable to directional cues are
moved from relatively higher to relatively lower frequencies.
7. A listening device according to claim 6 wherein said directional
cues are increased in magnitude.
8. A listening device according to claim 1 wherein the frequency
shaping unit is adapted to modify one or more selected time
frequency ranges to provide a directional frequency shaping of the
combined microphone signal in dependence of the direction of the
incoming sound signal.
9. A listening device according to claim 1 wherein the frequency
shaping unit is adapted to provide the directional frequency
shaping of the combined microphone signal in dependence of a users
hearing ability, e.g. frequency resolution and/or audiogram.
10. A listening device according to claim 1 wherein the directional
frequency shaping of the combined microphone signal comprises a
`roll off` of the directional microphone signal corresponding to a
rear direction of the user above a predefined ROLL-OFF-frequency,
e.g. above a frequency in the range from 1 kHz to 7 kHz.
11. A listening device according to claim 1 wherein the directional
frequency dependent shaping comprises inserting a peak or a notch
at a REAR-frequency in the resulting improved directional output
signal indicative of sound originating from a rear direction of the
user.
12. A listening device according to claim 10 wherein the
REAR-frequency is larger than or equal to 3 kHz, e.g. around 3 kHz
or around 4 kHz.
13. A listening device according to claim 1 wherein the ear-part
comprises a BTE-part adapted to be located behind an ear of a user,
the BTE-part comprising at least one microphone of the microphone
system.
14. A listening device according to claim 3 wherein the frequency
shaping-unit is adapted to provide that a frequency shaping is
applied to the combined microphone signal based on a decision of
whether a particular instance in time and frequency has its origin
from a particular direction, e.g. the front of the back of the
user.
15. A listening device according to claim 3 wherein the frequency
shaping-unit is adapted to implement a decision algorithm for
deciding whether or not or with which probability or weight a given
TF-range or unit is associated with a given spatial direction.
16. A listening device according to claim 15, wherein the decision
algorithm for each TF-range or unit is |CF|-|CR|.gtoreq..tau., in a
logarithmic expression, where |CF| and |CR| are the magnitudes of
the front and rear directional signals, respectively, and .tau. is
a directional bias constant.
17. A method of operating a listening device, the listening device
comprising an ear-part adapted for being worn in or at an ear of a
user, a front and rear direction being defined relative to a person
wearing the ear-part in an operational position, the method
comprising (a) providing at least two microphones signals, each
being an electrical representation of an input sound, (b) providing
a weighted sum of the at least two electrical microphone signals
resulting in at least two directional microphone signals having
maximum sensitivity in spatially different directions, e.g. in said
front and rear directions, and a combined microphone signal and (c)
modifying the combined microphone signal to indicate the
directional cues of input sounds originating from at least one of
said spatially different directions and providing an improved
directional output signal.
18. Use of a listening device according to claim 1.
19. A tangible computer-readable medium storing a computer program
comprising program code means for causing a data processing system
to perform at least some of the steps of the method of claim 17,
when said computer program is executed on the data processing
system.
20. A data processing system comprising a processor and program
code means for causing the processor to perform at least some of
the steps of the method of claim 17.
Description
TECHNICAL FIELD
[0001] The present invention relates to listening devices, e.g.
hearing aids, in particular to localization of sound sources
relative to a person wearing the listening device. The invention
relates specifically to a listening device comprising an ear-part
adapted for being worn in or at an ear of a user, a front and rear
direction being defined relative to a person wearing the ear-part
in an operational position.
[0002] The invention furthermore relates to a method of operating a
listening device, to its use, to a listening system, to a computer
readable medium and to a data processing system.
[0003] The invention may e.g. be useful in applications such as
listening devices, e.g. hearing instruments, head phones, headsets
or active ear plugs.
BACKGROUND ART
[0004] The following account of the prior art relates to one of the
areas of application of the present invention, hearing aids.
[0005] The localization cues for hearing impaired are often
degraded (due to the reduced hearing ability as well as due to the
configuration of a hearing aid worn by the hearing impaired),
meaning a degradation of the ability to decide from which direction
a given sound is received. This is annoying and can be dangerous,
e.g. in the traffic. The human localization of sound is related to
the difference in time of arrival, attenuation, etc. of a sound at
the two ears of a person and is e.g. dependent on the direction and
distance to the source of the sound, the form and size of the ears,
etc. These differences are modelled by the so-called Head-Related
Transfer functions (HRTFs). Further, the lack of spectral colouring
can make the perception of localization cues more difficult even
for monaural hearing aids (i.e. a system with a hearing instrument
at only one of the ears).
[0006] US 2007/0061026 A1 describes an audio processing system
comprising filters adapted for emulating `location-critical` parts
of HRTFs with the aim of creating or maintaining localization
related audio effects in portable devices, such as cell phones,
PDAs, MP3 players, etc.
[0007] EP 1 443 798 A2 deals with a hearing device with a
behind-the-ear microphone arrangement where beamforming provides
for substantially constant amplification independent of direction
of arrival of an acoustical signal at a predetermined frequency and
provides above such frequency directivity so as to reestablish a
head-related-transfer-function of the individual.
[0008] US 2007/230729 A1 deals with a hearing aid system comprising
a directional microphone system adapted for generating auditory
spatial cues. US 2009/0074197 A1 deals with a method of configuring
a frequency transposition scheme for transposing a set of received
frequencies of an audio signal received by a hearing aid worn by a
subject to a transposed set of frequencies.
[0009] A problem in particular with behind-the-ear (BTE) hearing
aids is that the microphones are placed above/behind the external
ear and thus this attenuation of sounds coming from behind
disappears. Front-back confusions are a common problem for hearing
impaired users of this kind of hearing aids.
DISCLOSURE OF INVENTION
[0010] However, it might be possible to introduce localization cues
for the hearing impaired, such as frequency-dependent attenuation
or direction-dependent peaks or notches. When comparing the
spectrally decomposed front and rear cardioids (see e.g. FIG. 2),
good front-rear estimation is obtained. Such a binary front-rear
decision can be used to enhance front-rear localization, by
applying different frequency shaping to the sound signal depending
on whether the signal impinges from the front or the rear.
[0011] An object of the present invention is to provide
localization cues for indicating a direction of origin of a sound
source.
[0012] Objects of the invention are achieved by the invention
described in the accompanying claims and as described in the
following.
A Listening Device:
[0013] An object of the invention is achieved by a listening device
comprising an ear-part adapted for being worn in or at an ear of a
user, a front and rear direction being defined relative to a person
wearing the ear-part in an operational position. The listening
device comprises (a) a microphone system comprising at least two
microphones each converting an input sound to an electrical
microphone signal, (b) a DIR-unit comprising a directionality
system for providing a weighted sum of the at least two electrical
microphone signals thereby providing at least two directional
microphone signals having maximum sensitivity in spatially
different directions and a combined microphone signal, and (c) a
frequency shaping-unit for modifying the combined microphone signal
to indicate directional cues of input sounds originating from at
least one of said spatially different directions and providing an
improved directional output signal.
[0014] This has the advantage of providing an alternative or an
addition to natural localization cues.
[0015] The term `indicate directional cues` is in the present
context taken to mean to `restore or enhance or replace` the
natural directional cues available for a normally hearing person
(without significant hearing impairment) under normal hearing
conditions (without extremely low or high sound pressure levels).
Directional cues in the combined microphone signal of a listening
device may be generated by the directional system of the DIR-unit
based on the two or more spatially dislocated microphones of the
microphone system. Such cues may be identified and transposed to a
frequency range appropriate for the user's wearing the listening
device. Directional cues in a listening device may alternatively or
additionally be artificially generated by the frequency shaping
unit based on information from the directional system regarding the
location of an acoustic source relative to a user wearing the
listening device. Such artificial cues may be adapted in magnitude
and frequency (e.g. frequency location and width) according to a
user's needs. In an embodiment, the at least two microphones are
located in the ear-part adapted for being worn in or at an ear of a
user. In an embodiment one of the at least two microphones is
located at an opposite ear of the user.
[0016] In the term `an improved directional output signal`,
`improved` is used in the sense that the output signal comprises
directional information that is aimed at providing an enhanced
perception by a user of the listening device.
[0017] In an embodiment, the `weighted sum of the at least two
electrical microphone signals` is taken to mean a weighted sum of a
complex representation of the at least two electrical microphone
signals. In an embodiment, the weighting factors are complex. The
`weighted sum of the at least two electrical microphone signals`
includes a linear combination of the at least two input signals
with a mutual delay between them. In an embodiment the microphone
system comprises two electrical microphone input signals TF1(f) and
TF2(f). A weighted sum of the two electrical microphone signals
providing e.g. a front directional signal CF, can thus be written
as CF(f)=TF1(f)w1F(f)+TF2(f)w2F(f), where f is frequency and
w1F(f), w2F(f) are (generally complex) weighting functions.
Correspondingly, a rear directional signal CR, can be written as
CR(f)=TF1(f)w1R(f)+TF2(f)w2R(f). In an embodiment, the weighting
functions can be adaptively determined (to achieve that the FRONT
and REAR directions are adaptively determined in relation to the
present acoustic sources).
[0018] In an embodiment, the listening device comprises a synthesis
unit comprising a time-frequency to time conversion arrangement
providing as an output a time dependent, improved directional
output signal comprising enhanced directional cues.
[0019] In an embodiment, the listening device comprises an output
transducer for presenting the improved directional output signal or
a signal derived there from as a stimulus adapted to be perceived
by a user as an output sound (e.g. an electro-acoustic transducer
(a receiver) of a hearing instrument or an output transducer (such
as a number of electrodes) of a cochlear implant or (such as a
vibrator) of a bone conducting hearing device).
[0020] A forward path of a listening device is defined as a signal
path from the input transducer (defining an input side) to an
output transducer (defining an output side).
[0021] In an embodiment, the listening device comprises an analogue
to digital (AD) converter unit providing said electrical microphone
signals as digitized electrical microphone signals.
[0022] In an embodiment, the listening device is adapted to be able
to perform signal processing (of the signal of the forward path
and/or of a control path influencing the signal of the forward
path) in separate frequency ranges or bands.
[0023] In an embodiment, the input side of the forward path of the
listening device comprises an AD-conversion unit for sampling an
analogue electric input signal with a sampling frequency f.sub.s
and providing as an output a digitized electric input signal
comprising digital time samples s.sub.n of the input signal
(amplitude) at consecutive points in time t.sub.n=n*(1/f.sub.s).
The duration in time of a sample is thus given by
T.sub.s=1/f.sub.s. In general, the sampling frequency is adapted to
the application (available bandwidth, power consumption, frequency
content of input signal, necessary accuracy, etc.). In an
embodiment, the sampling frequency f.sub.s is in the range from 8
kHz to 40 kHz, e.g. from 12 kHz to 24 kHz, e.g. around or equal to
16 kHz or 20 kHz.
[0024] In an embodiment, the listening device comprises a
TF-conversion unit for providing a time-frequency representation of
the at least two microphone signals, each signal representation
comprising corresponding complex or real values of the signal in
question in a particular time and frequency range. In an
embodiment, a signal of the forward path and/or a signal branched
off from the forward path is available in a time-frequency
representation, where a time representation of the signal exists
for each of the frequency bands constituting the frequency range
considered in the processing (from a minimum frequency f.sub.min to
a maximum frequency f.sub.max, e.g. from 10 Hz to 20 kHz, such as
from 20 Hz to 12 kHz). A `time-frequency region` may comprise one
or more adjacent frequency bands and one or more adjacent time
units.
[0025] In an embodiment, a number of consecutive samples s.sub.n
are arranged in time frames F.sub.m (m=1, 2, . . . ), each time
frame comprising a predefined number Q of digital time samples
s.sub.q (q=1, 2, . . . , Q) corresponding to a frame length in time
of L=Q/f.sub.s,=QT.sub.s, each time sample comprising a digitized
value s.sub.n (or s[n]) of the amplitude of the signal at a given
sampling time t.sub.n (or n). Alternatively, the time frames
F.sub.m may differ in length, e.g. according to a predefined
scheme.
[0026] In an embodiment, successive time frames (F.sub.m,
F.sub.m+1) have a predefined overlap of digital time samples. In
general, the overlap may comprise any number of samples .gtoreq.1.
In an embodiment, half of the Q samples of a frame are identical
from one frame F.sub.m to the next F.sub.m+1. In such embodiment,
F.sub.m={s.sub.m,1, s.sub.m,2, s.sub.m,(Q/2)-1, s.sub.m,Q/2,
s.sub.m,(Q/2)+1, s.sub.m,(Q/2)+2, . . . , s.sub.m,Q} and
F.sub.m+1={s.sub.m+1,1, s.sub.m+1, 2, . . . , s.sub.m+1,(Q/2)-1,
s.sub.m+1,Q/2, s.sub.m+1,(Q/2)+1, s.sub.m+1,(Q/2)+2, . . . ,
s.sub.m+1,Q}, where s.sub.m+1,1=s.sub.m,(Q/2)+1,
s.sub.m+1,2=s.sub.m,(Q/2)+2, . . . , s.sub.m+1,Q/2=s.sub.m,Q.
[0027] In an embodiment, the listening device is adapted to provide
a frequency spectrum of the signal in each time frame (m), a
time-frequency tile or unit comprising a (generally complex) value
of the signal in a particular time (m) and frequency (p) unit. In
an embodiment, only the real part (magnitude) of the signal is
considered, whereas the imaginary part (phase) is neglected. A
`time-frequency region` may comprise one or more adjacent
time-frequency units.
[0028] In an embodiment, the listening device comprises a
TF-conversion unit for providing a time-frequency representation of
a digitized electrical input signal and adapted to transform the
time frames on a frame by frame basis to provide corresponding
spectra of frequency samples, the time frequency representation
being constituted by TF-units each comprising a complex value
(magnitude and phase) or a real value (e.g. magnitude) of the input
signal at a particular unit in time and frequency. A unit in time
is in general defined by the length of a time frame minus its
overlap with its neighbouring time frame, e.g. corresponding to the
extension in time of the number of new time samples Q-N.sub.o of a
given time frame, where N.sub.o is the number of overlapping time
samples between a time frame and its previous time frame. In case
of no overlap, a time unit is equal to the frame length
L=Q/f.sub.s,=QT.sub.s. A unit in frequency is defined by the
frequency resolution of the time to frequency conversion unit. The
frequency resolution may vary over the frequency range considered,
e.g. to have an increased resolution at relatively lower
frequencies compared to at relatively higher frequencies.
[0029] In an embodiment, the listening device is adapted to provide
that the spatially different directions are said front and rear
directions.
[0030] In an embodiment, the DIR-unit is adapted to detect from
which of the spatially different directions a particular time
frequency region or TF-unit originates. This can be achieved in
various different ways as e.g. described in U.S. Pat. No. 5,473,701
or in WO 99/09786 A1.
[0031] In an embodiment, the spatially different directions are
adaptively determined, cf. e.g. U.S. Pat. No. 5,473,701 or EP 1 579
728 B1.
[0032] In an embodiment, the frequency shaping unit is adapted to
apply directional cues, which would naturally occur in a given time
frequency range, in a relatively lower frequency range. In an
embodiment, the frequency shaping-(FS-) unit is adapted to apply
directional cues of a given time frame, occurring naturally in a
given frequency region or unit, in relatively lower frequency
regions or frequency units. In the present context, a `relatively
lower frequency region or frequency unit` compared to a given
frequency region or unit (at a given time) is taken to mean a
frequency region or unit representing a frequency f.sub.x that is
lower than the frequency f.sub.p at the given time or time unit
(i.e. has a lower index x than the frequency f.sub.p (x<p) in
the framework of FIG. 3).
[0033] In an embodiment, the applied directional cues are increased
in magnitude compared to naturally occurring directional cues. In
an embodiment, the increase is in the range from 3 dB to 30 dB,
e.g. around 10 dB or around 20 dB.
[0034] In an embodiment, differences in the microphone signals from
different directions (e.g. front and rear) attributable to
directional cues are moved from the naturally occurring, relatively
higher, frequencies to relatively lower frequencies or frequency
units. The microphones may be located at the same ear or,
alternatively, at opposite ears of a user.
[0035] In an embodiment, the directional cues (e.g. a number Z of
notches located at different frequencies, f.sub.N1, f.sub.N2,
f.sub.Nz) are modeled and applied at relatively lower frequencies
than the naturally occurring frequencies. In an embodiment, the
notches inserted at relatively lower frequencies have the same
frequency spacing as the original ones. In an embodiment, the
notches inserted at relatively lower frequencies have a compressed
frequency spacing. This has the advantage of allowing a user to
perceive the cues, even while having a hearing impairment at the
frequencies of the directional cues. In an embodiment, the
directional cues are increased in magnitude (compared to their
natural values). In an embodiment, the magnitude of a notch is in
the range from 3 dB to 30 dB, e.g. 3 dB to 5 dB or 10 dB to 30
dB.
[0036] In an embodiment, the notches are wider in frequency than
corresponding naturally occurring notches. In an embodiment, the
width in frequency and/or magnitude of a notch applied as a
directional cue is determined depending on a user's hearing
ability, e.g. frequency resolution or audiogram. In an embodiment,
the notches (or peaks) extend over more than one frequency band in
width. In an embodiment, the notches (or peaks) are up to 500 Hz in
width, such as up to 1 kHz in width, such as such as up to 1.5 kHz
or 2 kHz or 3 kHz in width. In an embodiment, the width of a peak
or notch is adjusted during fitting of a listening device to a
particular user's needs.
[0037] In general the frequency shaping can be performed on any
weighted (e.g. linear) combination of the input electrical
microphone signals, here termed `the combined microphone signal`
(e.g. TF1(f)w1c(f)+TF2(f)w2c(f)). The resulting signal after the
frequency shaping is here termed the `improved directional signal`
(even if the combined microphone signal is (chosen to be) an
omni-directional signal, `directional` here relating to the
directional cues). In an embodiment, the signal wherein the
frequency shaping is performed is a signal, which is intended for
being presented to a user (or chosen for further processing with
the aim of later presentation to a user). In an embodiment, the
frequency shaping is performed on one of the input microphone
signals or on one of the directional microphone signals provided by
the DIR-unit or on weighted combinations thereof. In an embodiment,
the FS-unit is adapted to modify one or more selected TF-units or
ranges to provide a directional frequency shaping of the combined
microphone signal in dependence of the direction of the incoming
sound signal.
[0038] In an embodiment, the FS-unit is adapted to provide that
different frequency shaping is applied to the combined microphone
signal based on a (binary or non-binary) decision of whether a
particular instance in time and frequency (a TF-bin or unit) has
its origin from a particular direction, e.g. the front of the back
of the user. This has the advantage of restoring or enhancing the
natural front-back cues. In an embodiment, the FS-unit is adapted
to implement a decision algorithm for deciding whether or not (or
with which probability or weight) a given TF-range or unit is
associated with a given spatial direction. In an embodiment, the
decision algorithm (for each TF-range or unit) is
|CF|-|CR|.gtoreq..tau., in a logarithmic expression, where |CF| and
|CR| are the magnitudes of the front and rear directional signals,
respectively, and .tau. is a (directional) bias constant. The
algorithm can e.g. be interpreted in a binary fashion to indicate
that the signal component of that TF-range or unit is assumed to
originate from a FRONT direction, if the expression is TRUE, and
the signal is assumed to originate from a REAR direction, if the
expression is FALSE. Alternatively, a continuous interpretation can
be applied, e.g. in that the (possibly normalized) value of the
expression |CF|-|CR|-.tau. is used as a measure of the probability
or weight with which the TF-range or unit in question belongs to a
given spatial direction (positive values indicating FRONT and
negative values indicating REAR).
[0039] In an embodiment, the FS-unit is adapted to provide the
directional frequency shaping of the combined microphone signal in
dependence of a users hearing ability, e.g. an audiogram or
depending on the user's frequency resolution. Preferably, the
directional cues are located at frequencies, which are adapted to a
user's hearing ability, e.g. located at frequencies where the
user's hearing ability is acceptable. In an embodiment, the
specific directional frequency shaping (representing directional
cues) is determined during fitting of a listening device to a
particular user's needs.
[0040] In an embodiment, the directional frequency shaping of the
combined microphone signal comprises a `roll off` corresponding to
a specific direction, e.g. a rear direction, of the user above a
predefined ROLL-OFF-frequency f.sub.roll, e.g. above 1 kHz, such as
above 1.5 kHz, such as above 2 kHz, such as above 3 kHz, such as
above 4 kHz, such as above 5 kHz, such as above 6 kHz, such as
above 7 kHz, such as above 8 kHz. In an embodiment, the predefined
roll off frequency is adapted to a user's hearing ability, to
ensure sufficient hearing ability at the roll off frequency. The
term `roll off` is in the present context taken to mean `decrease
with increasing frequency`, e.g. linearly on a logarithmic
scale.
[0041] In an embodiment, the directional frequency dependent
shaping comprises inserting a peak or a notch at a REAR-frequency
in the resulting improved directional output signal indicative of
sound originating from a rear direction of the user. In an
embodiment, the REAR-frequency is larger than or equal to 3 kHz,
e.g. around 3 kHz or around 4 kHz. In an embodiment, the
directional frequency dependent shaping is ONLY performed for
sounds originating from a rear direction of the user. In an
embodiment, directional frequency dependent shaping comprises
inserting a peak or a notch at a FRONT-frequency in the resulting
improved directional output signal indicative of sound originating
from a front direction of the user. In an embodiment, the
FRONT-frequency is larger than or equal to 3 kHz, e.g. around 3 kHz
or around 4 kHz.
[0042] In an embodiment, the peaks or notches deviate from a
starting level by a predefined amount, e.g. by 3-30 dB, e.g. by 10
dB.
[0043] In an embodiment, the peaks or notches are inserted in a
range from 1 kHz, to 5 kHz.
[0044] In an embodiment, the ear-part comprises a BTE-part adapted
to be located behind an ear of a user, the BTE-part comprising at
least one microphone of the microphone system. In an embodiment,
the ear-part comprises the at least two microphones of the
microphone system. In an embodiment, the BTE-part comprises the at
least two microphones of the microphone system.
[0045] In an embodiment, the listening device comprises a hearing
instrument adapted for being worn at or in an ear and providing a
frequency dependent gain of an input sound. In an embodiment, the
hearing instrument is adapted for being worn by a user at or in an
ear. In an embodiment, the hearing instrument comprises a behind
the ear (BTE) part adapted for being located behind an ear of the
user, wherein at least one microphone (e.g. two microphones) of the
microphone system is located in the BTE part. In an embodiment, the
hearing instrument comprises an in the ear (ITE) part adapted for
being located fully or partially in the ear canal of the user. In
an embodiment, at least one microphone of the microphone system is
located in the ITE part. In an embodiment, the hearing instrument
comprises an input transducer (e.g. a microphone) for converting an
input sound to en electric input signal, a signal processing unit
for processing the input signal according to a user's needs and
providing a processed output signal and an output transducer (e.g.
a receiver) for converting the processed output signal to an output
sound. In an embodiment, the hearing instrument comprises a noise
reduction system (e.g. an anti-feedback system). In an embodiment,
the hearing instrument comprises a compression system.
[0046] In an embodiment, the listening device is a low power,
portable device comprising its own energy source, e.g. a
battery.
[0047] In an embodiment, the listening device comprises an
electrical interface to another device allowing reception (or
interchange) of data (e.g. directional cues) from the other device
via a wired connection. The listening device may, however, in a
preferred embodiment comprise a wireless interface adapted for
allowing a wireless link to be established to another device, e.g.
to a device comprising a microphone contributing to the
localization of audio signals (e.g. a microphone of the microphone
system). In an embodiment, the other device is a physically
separate device (from the listening device, e.g. another body-worn
device). In an embodiment, the microphone signal from the other
device (or a part thereof, e.g. one or more selected frequency
ranges or bands or a signal related to localization cues derived
from the microphone signal in question) is transmitted to the
listening device via a wired or wireless connection. In an
embodiment, the other device is the opposite hearing instrument of
a binaural fitting. In an embodiment, the other device is an audio
selection device adapted to receive a number of audio signals and
to transmit one of them to the listening device in question. In an
embodiment, localization cues derived from a microphone of another
device is transmitted to the listening device via an intermediate
device, e.g. an audio selection device. In an embodiment, a
listening device is able to distinguish between 4 spatially
different directions, e.g. FRONT, REAR, LEFT and RIGHT.
Alternatively, a directional microphone system comprising more than
two microphones, e.g. 3 or 4 or more microphones can be used to
generate more than 2 directional microphone signals. This has the
advantage that the space around a wearer of the listening device
can be divided into e.g. 4 quadrants, allowing different
directional cues to be applied indicating signals originating from
e.g. LEFT, REAR, RIGHT directions relative to a user, which greatly
enhances the orientation ability of a wearer relative to acoustic
sources. In an embodiment, the applied directional cues comprise
peaks or notches or combinations of peaks and notches, e.g. of
different frequency, and/or magnitude, and/or width to indicate the
different directions.
[0048] In an embodiment, the listening device comprises an active
ear plug adapted for protecting a person's hearing against
excessive sound pressure levels. In an embodiment, the listening
device comprises a headset and/or an earphone.
A Listening System:
[0049] A listening system comprising a pair of listening devices as
described above, in the detailed description of `mode(s) for
carrying out the invention` and in the claims is furthermore
provided. In an embodiment, the listening system comprises a pair
of hearing instruments adapted for aiding in compensating a persons
hearing impairment on both ears. In an embodiment, the two
listening devices are adapted to be able to exchange data
(including microphone signals or parts thereof, e.g. one or more
selected frequency ranges thereof), preferably via a wireless
connection, e.g. via a third, intermediate, device, such as an
audio selection device. This has the advantage that location
related information (localization or directional cues) can be
better extracted (due to the spatial difference of the input
signals picked up by the two listening devices).
A Method:
[0050] A method of operating a listening device, the listening
device comprising an ear-part adapted for being worn in or at an
ear of a user, a front and rear direction being defined relative to
a person wearing the ear-part in an operational position is
furthermore provided by the present invention. The method comprises
(a) providing at least two microphones signals, each being an
electrical representation of an input sound, (b) providing a
weighted sum of the at least two electrical microphone signals
resulting in at least two directional microphone signals having
maximum sensitivity in spatially different directions, e.g. in said
front and rear directions, and a combined microphone signal and (c)
modifying the combined microphone signal to indicate the
directional cues of input sounds originating from at least one of
said spatially different directions and providing an improved
directional output signal.
[0051] It is intended that the structural features of the listening
device described above, in the detailed description of `mode(s) for
carrying out the invention` and in the claims can be combined with
the method, when appropriately substituted by a corresponding
process. Embodiments of the method have the same advantages as the
corresponding listening device.
[0052] In an embodiment, the method comprises providing the at
least two electrical microphone signals in a digitized form and
providing a time-frequency representation of said digitized
electrical microphone signals, said time frequency representation
being constituted by TF-units each comprising a complex or real
value of the microphone signal in question at a particular unit in
time and frequency. One or more of the digitized electrical
microphone signals may originate from a device separate from the
listening device in question.
Use of a Listening Device:
[0053] Use of a listening device as described above, in the
detailed description of `mode(s) for carrying out the invention`
and in the claims is moreover provided by the present invention. In
particular embodiments, use in a hearing instrument, in an active
ear plug or in a pair of ear phones or in a head set is provided.
In an embodiment, the listening device is used in a gaming
situation to enhance localization cues in connection with a
computer game.
A Computer-Readable Medium:
[0054] A tangible computer-readable medium storing a computer
program comprising program code means for causing a data processing
system to perform at least some of the steps (e.g. at least steps
(b) and (c)) of the method described above, in the detailed
description of `mode(s) for carrying out the invention` and in the
claims, when said computer program is executed on the data
processing system is furthermore provided by the present
invention.
A Data Processing System:
[0055] A data processing system comprising a processor and program
code means for causing the processor to perform at least some of
the steps (e.g. at least steps (b) and (c)) of the method described
above, in the detailed description of `mode(s) for carrying out the
invention` and in the claims is furthermore provided by the present
invention.
[0056] Further objects of the invention are achieved by the
embodiments defined in the dependent claims and in the detailed
description of the invention.
[0057] As used herein, the singular forms "a," "an," and "the" are
intended to include the plural forms as well (i.e. to have the
meaning "at least one"), unless expressly stated otherwise. It will
be further understood that the terms "includes," "comprises,"
"including," and/or "comprising," when used in this specification,
specify the presence of stated features, integers, steps,
operations, elements, and/or components, but do not preclude the
presence or addition of one or more other features, integers,
steps, operations, elements, components, and/or groups thereof. It
will be understood that when an element is referred to as being
"connected" or "coupled" to another element, it can be directly
connected or coupled to the other element or intervening elements
maybe present, unless expressly stated otherwise. Furthermore,
"connected" or "coupled" as used herein may include wirelessly
connected or coupled. As used herein, the term "and/or" includes
any and all combinations of one or more of the associated listed
items. The steps of any method disclosed herein do not have to be
performed in the exact order disclosed, unless expressly stated
otherwise.
BRIEF DESCRIPTION OF DRAWINGS
[0058] The invention will be explained more fully below in
connection with a preferred embodiment and with reference to the
drawings in which:
[0059] FIG. 1 shows directional transfer functions for the right
ears of two subjects with small (first and third panels) and large
pinnae (second and fourth panels), respectively (from
[Middlebrooks, 1999]),
[0060] FIG. 2 shows parts of a listening device according to an
embodiment of the invention,
[0061] FIG. 3 schematically shows a time-frequency mapping of a
time dependent input signal,
[0062] FIG. 4 shows a listening device according to an embodiment
of the invention,
[0063] FIG. 5 schematically illustrates an example of FRONT (FIG.
5a) and REAR directional cues (FIG. 5b) and a directional
time-frequency representation of an input signal (FIG. 5c)
according to an embodiment of the invention,
[0064] FIG. 6 shows a time frequency representation of a FRONT and
REAR microphone signal, CF and CR, respectively, (FIG. 6a), a
differential microphone signal CF-CR (FIG. 6b), and a binary
time-frequency mask representation of the differential microphone
signal (FIG. 6c),
[0065] FIG. 7 shows various exemplary directional cues (linear
scale) for introduction in FRONT and REAR microphone signals
according to an embodiment of the invention, FIG. 7a illustrating a
decreasing gain beyond a roll-off frequency for a signal
originating from a REAR direction, and FIGS. 7b and 7c directional
cues in the form of peaks or notches at predefined frequencies in
the FRONT and/or REAR signals, respectively, and
[0066] FIG. 8 shows embodiments of a listening device comprising an
ear-part adapted for being worn at an ear of a user, FIG. 8a
comprising a BTE-part comprising two microphones, FIG. 8b
comprising a BTE-part comprising two microphones and a separate,
auxiliary device comprising at least a third microphone.
[0067] The figures are schematic and simplified for clarity, and
they just show details which are essential to the understanding of
the invention, while other details are left out.
[0068] Further scope of applicability of the present invention will
become apparent from the detailed description given hereinafter.
However, it should be understood that the detailed description and
specific examples, while indicating preferred embodiments of the
invention, are given by way of illustration only, since various
changes and modifications within the spirit and scope of the
invention will become apparent to those skilled in the art from
this detailed description.
MODE(S) FOR CARRYING OUT THE INVENTION
[0069] The shape of the external ears influences the attenuation of
sounds coming from behind. The attenuation is frequency dependent
and is typically larger at higher frequencies.
[0070] A problem in particular with behind-the-ear (BTE) hearing
aids is that the microphones are placed above/behind the external
ear and thus this attenuation of sounds coming from behind
disappears (cf. e.g. FIG. 8). Front-back confusions are a common
problem for hearing impaired users of this kind of hearing aids. It
is proposed to compensate for that by applying different frequency
shaping based on a decision (possibly binary) of whether a
particular instance in time and frequency (a TF-bin or unit) has
its origin from the front of the back of the user, thus restoring
or enhancing the natural front-back cues.
[0071] The terms `front-back` and `front-rear` are used
interchangeably with no intended difference in meaning.
[0072] A further possibility is to not just compensate for the BTE
placement, but to further increase the front-back difference, e.g.
by increasing the front-back difference further down in
frequencies. An enhanced front-back difference would correspond to
increasing the size of the listener's pinna (like when people place
their hands behind the ear in order to focus attention on the
speaker in front of them). This suggestion could be used with any
hearing aid style. It is useful in particular for hearing impaired
persons because they often loose high-frequency hearing, and the
normal-sized pinna has a frequency shaping effect that is confined
mainly to high frequencies.
[0073] The subject, often referred to as `the cone of confusion`,
is e.g. discussed in [Blauert et al., 1997], page 179.
[0074] FIG. 1 shows directional transfer functions for the right
ears of two subjects with small (first and third panels) and large
pinnae (second and fourth panels), respectively (from
[Middlebrooks, 1999]). Left panels show responses for different
elevation angles along the frontal midline, and right panels show
responses for different elevation angles along the rear midline.
0.degree. corresponds to a source at the same horizontal plane as
the ears, and positive angles to positions above that plane. The
transfer functions are similar among subjects, but might be offset
in frequency due to different physical dimensions. If one looks at
typical head-related transfer functions, there is a clear spectral
shape difference between front (FIG. 1, left panels) and back (FIG.
1, right panels). The difference is clearest at the median plane
(0.degree. elevation), and mainly confined to frequencies above 5
kHz. The preferred implementation would try to restore these
high-frequency spectral cues. Such restoration could e.g. be
established taking account of a user's hearing ability. Typically a
restoration at lower frequencies, where a user has better hearing
ability, is preferable. Depending on the user's hearing profile, an
amplification of the restored directional information can be
performed.
[0075] Alternatively or additionally, new front-back cues can be
introduced. E.g. if the sound impinges from the front, a notch (or
a peak) at 3 kHz can be applied, and/or if the sound arrives from
behind, a notch (or a peak) at 4 kHz can be applied. When exposed
to such a direction-dependent frequency shaping for some time, the
hearing impaired will be able to learn to distinguish between
sounds impinging from the front and the rear direction. This
artificial frequency dependent shaping can also be made dependent
on the particular user's hearing ability, e.g. frequency resolution
and/or the shape of the audiogram of the user. Artificial cues can
for instance be used for users with virtually no residual
high-frequency hearing, and independent of device style (i.e. NOT
confined to BTE-type devices).
[0076] An example of such a directional cue-introducing system is
illustrated in FIG. 2. FIG. 2 shows parts of a listening device
according to an embodiment of the invention. Electrical signals IN1
and IN2 representing sound inputs as e.g. picked up by two
microphones are fed to each their Analysis unit for providing a
time to frequency conversion (e.g. as implemented by a filter bank
or a Fourier transformation unit). The outputs of the Analysis
units comprise a time-frequency representation of the input signals
IN1 and IN2, respectively. In the directional unit termed C.sub.F,
C.sub.R comparison in FIG. 2, directional signals CF and CR are
created, each being a weighted combination of the (time frequency
representation of the) input signals IN1 and IN2 and representing
outputs of a front aiming and rear aiming microphone sensitivity
characteristic (cardioid), respectively. By comparing a front and a
rear cardioid, it is possible to determine if a sound impinges from
the front or from the rear direction. In practice, the time
frequency representations of signals CF and CR are compared and a
differential time frequency (TF) map is generated based on a
predefined criterion. Each TF-map comprises the magnitude and/or
phase of CF (or CR) at different instances in time and frequency.
Preferably, a time frequency map comprises TF-units (m,p) covering
the time and frequency ranges considered for the application in
question. In the following, the respective TF-maps of CF and CR are
assumed to comprise only the magnitudes |.cndot.| of the signals.
The output of the directional unit termed C.sub.F, C.sub.R
comparison unit in FIG. 2, are the TF maps of signals CF and CR
comprising respective magnitudes (or gains) of CF and CR, which are
fed to the Binary decision unit comprising an algorithm for
deciding the direction of origin of a given TF-range or unit.
[0077] One algorithm for a given TF-range or unit can e.g. be IF
(|CF|-|CR|.gtoreq..tau., in a logarithmic expression), the signal
component of that range or unit is assumed to originate from a
FRONT direction; otherwise, the signal is assumed to originate from
a REAR direction. In general the real constant .tau. in dB
determines the focus of the application (e.g. the polar angle used
to distinguish between FRONT and REAR), positive values of .tau.
[dB] indicating a focus in the FRONT direction, negative values of
.tau. [dB] indicating a focus in the REAR direction. In an
embodiment, the threshold value .tau. equals 0 [dB]. Values
different from 0 [dB] can e.g. be founded on one of the signals
being better estimated or more accurate than another. Such a
decision can in general be gradual (e.g. comprising several steps
between FRONT and REAR). In an embodiment, the decision is binary
(as indicated by the Binary decision unit of FIG. 2). A
corresponding algorithm can e.g. be IF
(|CF(m,p)|-|CR(m,p)|.gtoreq..tau.), BTF(m,p)=1; otherwise
BTF(m,p)=0. In an embodiment, the threshold value .tau. equals 0
[dB]. The output of the Binary decision unit is such binary BTF-map
holding a binary representation of the origin of each TF-unit. The
output is, e.g. together with the TF maps of signals CF and/or CR
and/or another weighted combination of the electric microphone
signals, fed to a frequency shaping unit (cf. Front-rear-dependent
frequency shaping unit in FIG. 2). In the frequency shaping unit, a
localization cue is introduced and/or re-established by applying a
certain frequency-shaping when the sound impinges from the front
and/or another frequency-shaping when the sound impinges from the
rear direction. In general, a map of gains (magnitudes) of the
chosen signal (a directional or omni-directional signal) to be used
as a basis for further processing (e.g. presentation to a user) can
be multiplied by a chosen cue gain map. A FRONT cue gain map
GC.sub.front(G.sub.f1, G.sub.f2, . . . , G.sub.fP) can e.g. be
multiplied on the BTF.sub.front(m,p) map to provide a
GC.sub.front(m,p) map and/or a REAR cue gain map
GC.sub.rear(G.sub.r1, G.sub.r2, . . . , G.sub.rP) can e.g. be
multiplied on the BTF.sub.rear(m,p) map
(BTF.sub.rear(m,p)=1(m,p)-BTF.sub.front(m,p)) to provide a
GC.sub.rear(m,p) map. The GC.sub.front(m,p) map is e.g. generated
by vector multiplying the GC.sub.front vector with each column of
the BTF.sub.front(m,p) map. If, e.g., we want to introduce a rear
cue in a resulting directional microphone signal (comprising a
weighted sum of the input microphone signals), the GC.sub.rear(m,p)
map is multiplied on the G.sub.dir(m,p) map of the directional
microphone signal providing an improved directional output signal
G.sub.imp-dir(m,p), where
G.sub.imp-dir(m,p)=G.sub.dir(m,p)GC.sub.rear(m,p). In an
embodiment, the directional microphone signal has a preferred (e.g.
front aiming) directional sensitivity. In an embodiment, the
directional microphone signal is an omni-directional signal
comprising the sum of the individual input microphone signals (here
IN1(f) and IN2(f)). In the embodiment of FIG. 2, the improved
directional output signal is the output of the Front-rear-dependent
frequency shaping unit. This output signal is fed to a Synthesis
unit comprising a time-frequency to time conversion arrangement
providing as an output a time dependent, improved directional
output signal comprising enhanced directional cues. The improved
directional output signal can be presented to a user via an output
transducer or be fed to a signal processing unit for further
processing (e.g. for applying a frequency dependent gain according
to a user's hearing profile), cf. e.g. FIG. 4.
[0078] FIG. 3 shows a time-frequency mapping of a time dependent
input signal. An AD-conversion unit samples an analogue electric
input signal with a sample frequency f.sub.s and provides a
digitized electrical signal x.sub.n. The digitized electrical
signal x.sub.n is e.g. arranged in time frames each comprising a
predefined number Q of digital time samples x.sub.q (q=1, 2, . . .
, Q), corresponding to a frame length in time of L=Q/f.sub.s, where
f.sub.s is the sampling frequency of the AD-conversion unit. A
number of consecutive time frames are stored in a memory. A
time-frequency representation of the digitized signal is provided
by transforming the stored time frames on a frame by frame basis to
generate corresponding spectra of frequency samples, the time
frequency representation being constituted by TF-units (cf.
TF-unit(m,p) in FIG. 3) each comprising a generally complex value
of the input signal at a particular unit in time .DELTA.t and
frequency .DELTA.f. FIG. 3 shows a M.times.P map comprising a
number of M time units .DELTA.t.sub.m, m=1, 2, . . . , M, each
comprising a number of P frequency units .DELTA.f.sub.p, p=1, 2, .
. . , P. In general, the complex value of each TF-unit comprises
real (magnitude) and imaginary parts (phase angle) of the input
signal in the particular time and frequency unit (.DELTA.t.sub.m,
.DELTA.f.sub.p). In an embodiment, only the magnitude of the signal
is considered.
[0079] FIG. 4 shows a listening device according to an embodiment
of the invention. The listening device comprises a microphone
system comprising two (e.g. omni-directional) microphones receiving
input sound signals S1 and S2, respectively. The microphones
convert the input sound signals S1 and S2 to electric microphone
signals IN1 and IN2, respectively. The electric microphone signals
IN1 and IN2 are fed to respective time to time-frequency conversion
units A1, A2. In the present embodiment, time to time-frequency
conversion units A1, A2 provide time-frequency representations TF1,
TF2, respectively of the electric microphone signals IN1 and IN2
(cf. e.g. FIG. 3). The time-frequency representations TF1, TF2, are
fed to a directionality unit DIR comprising a directionality system
for providing a weighted sum of the at least two electrical
microphone signals resulting in at least two directional microphone
signals CF, CR having maximum sensitivity in spatially different
directions, here FRONT and REAR directions relative to a user's
face. The (time-frequency representations of the) output signals
CF, CR of the DIR-unit are fed to a decision unit DEC for
estimating on a unit by unit basis whether a particular time
frequency component has its origin from a mainly FRONT or mainly
REAR direction. In the present embodiment, the time-frequency
representations of signals CF and CR are compared and a
differential time frequency (TF) map FRM (e.g. a binary map, BTF)
is generated based on a predefined criterion. The output (signal or
TF-map FRM) of the decision unit DEC is fed to a frequency
shaping-unit FS for to generate the directional cues of input
sounds originating from said spatially different directions (here
FRONT and REAR) and providing an output signal GC comprising the
introduced gain cues (e.g. FRONT gain cues and/or REAR gain cues
applied to the differential time frequency (TF) map FRM). The
output signal(s) GC from the frequency shaping unit FS are fed to a
multiplication unit X (alternatively included in the FS-unit),
wherein the output signal(s) GC comprising the introduced gain cues
is/are multiplied to the corresponding directional signal WIN
comprising a weighted sum of the microphone signals (or rather of
TF-representation thereof), here extracted from the D/R-unit:
WIN(f)=TF1(f)w1(f)+TF2(f)w2(f), where f is frequency and w1(f),
w2(f) are weighting functions, which in an embodiment can be
adaptively determined (to achieve that the FRONT and REAR
directions are adaptively determined in relation to the present
acoustic sources). The resulting output WINXGC of the
multiplication unit X represents an improved directional output
signal comprising new, improved and/or reestablished directional
cues. In the embodiment of FIG. 4, this signal is fed to a signal
processing unit G for further processing the improved directional
output signal WINXGC, e.g. introducing further noise reduction,
compression and/or anti feedback algorithms and/or for providing a
frequency dependent gain according to a particular user's needs.
The output GOUT of the signal processing unit G is fed to a
synthesis unit S for converting the time frequency representation
of the output GOUT to a time domain output signal OUT, which is fed
to a receiver for being presented to a user as an output sound. In
embodiments, one or more of the processing algorithms are
introduced before the introduction of localization cues.
[0080] In the embodiment of FIG. 4, the order of the time to
time-frequency conversion units A1, A2 and the directionality unit
DIR may alternatively be switched, so that directional signals are
created before a time to time-frequency conversion is
performed.
[0081] FIG. 5 illustrates an example of FRONT (FIG. 5a) and REAR
directional cues (FIG. 5b) and a directional time-frequency
representation of an input signal (FIG. 5c) according to an
embodiment of the invention. An artificial directional cue in the
form of a forced attenuation of a directional signal originating
from the REAR can preferably be introduced. In FIGS. 5a and 5b,
corresponding exemplary directional gain cues, i.e. gain vs.
frequency, are illustrated. FIG. 5a shows a flat FRONT gain cue
graph GC.sub.from(f) [dB]=0 dB, f being frequency (here illustrated
by splitting the frequency range considered f.sub.min-f.sub.max in
12 frequency bands, f.sub.1, f.sub.2, . . . , f.sub.12). A
corresponding FRONT cue gain vector GC.sub.front(p)=1 (linear),
p=1, 2, . . . , 12 is shown. FIG. 5b shows a REAR gain cue graph
GC.sub.rear(f) [dB] having a flat part below a roll-off frequency
f.sub.roll and a roll-off in the form of an increasing attenuation
(here a linearly increasing attenuation (or decreasing gain) on a
logarithmic scale [dB]) at frequencies larger than f.sub.roll. The
roll-off frequency is preferably adapted to a user's hearing
profile to ensure that the decreasing gain beyond f.sub.roll
constituting a REAR gain cue is perceivable to the user. A
corresponding REAR cue gain vector GC.sub.rear(P)=1, p=1, 2, . . .
, 6, GC.sub.rear(P)=1/2.sup.(p-6), p=7, 8, . . . , 12 is shown
(linear). Here, the roll-off frequency f.sub.roll=f.sub.6. FIG. 5c
shows a time frequency map based on a FRONT and REAR directional
signal, F or R in a specific TF-unit indicating that the signal
component of the TF-unit originates from a FRONT or REAR direction,
respectively, relative to a user as determined by a decision
algorithm based on the corresponding FRONT and REAR directional
signals. `F` and `R` may e.g. be replaced by a 1 and 0,
respectively, or by a 0 and 1, respectively, as the case may be.
The frequency range considered may comprise a smaller or larger
amount of frequency ranges or bands than 12, e.g. 8 or 16 or 32 or
64 or more. The minimum frequency f.sub.min considered may e.g. be
in the range from 10 to 30 Hz, e.g. 20 Hz. The maximum frequency
f.sub.max considered may e.g. be in the range from 6 kHz to 30 kHz,
e.g. 8 kHz or 12 kHz or 16 kHz or 20 kHz. The roll-off frequency
f.sub.roll may e.g. be in the range from 2 kHz to 8 kHz, e.g.
around 4 kHz. The gain reduction may e.g. be in the range from 10
dB/decade to 40 dB/decade, e.g. around 20 dB/decade.
[0082] FIG. 6 shows a time frequency representation of a FRONT and
REAR microphone signal, CF and CR, respectively, (FIG. 6a), a
differential microphone signal CF-CR (FIG. 6b), and a binary
time-frequency mask representation of the differential microphone
signal (FIG. 6c). The frequency range considered is divided in 8
frequency ranges or bands, each comprising a single frequency
f.sub.p, p=1, 2, . . . , 8. Frequency spectra f.sub.p determined at
a number of consecutive time instances t.sub.m, m=1, 2, . . . , 12
constitute a time-frequency map TF(m,p), each TF-unit(m,p)
comprising a magnitude value of the signal (in an arbitrary scale)
at that frequency p and time unit m. FIG. 6a shows exemplary
corresponding time-frequency maps TF.sub.front(m,p) and
TF.sub.rear(m,p), each mapping magnitudes |CF(m,p)| and |CR(m,p)|,
e.g. in a logarithmic scale [dB]. A sound signal from a FRONT
direction predominates in time units m=1-6, whereas a sound signal
from a REAR direction predominates in time units m=8-12 as
illustrated in the TF-map of the differential signal |CF|-|CR| in
FIG. 6b. A binary TF-map, BTM, of the differential signal |CR|-|CF|
defined by the criterion IF |CR(m,p)|-|CF(m,p)|>0, BTM(m,p)=1,
ELSE BTM(m,p)=0, m=1, 2, . . . , 12, p=1, 0.2, . . . , 8 is shown
in FIG. 6c. As it appears, in the shown time frames, the sound
signal sources are predominantly FRONT in the first 6 time frames
and predominantly originating from the REAR in the last 6 time
frames. There are however, a few TF-units in the first 6 time
frames that originate from the REAR and a few TF-units in the last
6 time frames that originate from the FRONT. This represents one of
the strengths of the TF-masking method that the processing can be
performed on each individual TF-unit.
[0083] FIG. 7 shows various exemplary directional cues (linear
scale) for introduction in FRONT and REAR microphone signals
according to an embodiment of the invention, FIG. 7a illustrating a
decreasing gain beyond a roll-off frequency for a signal
originating from a REAR direction, and FIGS. 7b and 7c directional
cues in the form of peaks or notches at predefined frequencies in
the FRONT and/or REAR signals, respectively. The frequency range
considered is divided in 8 frequency ranges or bands, each
comprising a single frequency f.sub.p, p=1, 2, . . . , 8. FIG. 7a
illustrates a flat unity gain for signals from a FRONT direction
and a flat unity gain up to roll-off frequency f.sub.roll=f.sub.4
with a decreasing gain above the roll-off frequency (similar to
FIG. 5a, 5b). FIG. 7b shows a flat unity gain for signals from a
FRONT direction and a REAR directional cue in the form of a notch
at a frequency f.sub.7. FIG. 7c shows a FRONT directional cue in
the form of a peak at a frequency f.sub.5 and a REAR directional
cue in the form of a notch at a frequency f.sub.7. Other
directional cues may be envisaged, e.g. comprising more than one
peak or notch at different frequencies or comprising a mixture of
one or more peaks and one or more notches at different frequencies.
In an embodiment, natural cues as e.g. illustrated in FIG. 1 are
modelled, e.g. as a number of notches (e.g. 3-5) at frequencies
above 5 kHz. In an embodiment, the magnitudes in dB of the notches
are around 20 dB. In an embodiment, magnitude in dB of the notches
is increased compared to their natural values, e.g. to more than 30
dB, e.g. in dependence of a user's hearing impairment at the
frequencies in question. In an embodiment, the notches (or peaks)
are `relocated` to lower frequencies than their natural appearance
(e.g. depending on the user's hearing impairment at the frequencies
in question). In an embodiment, the notches (or peaks) are wider
than the naturally occurring directional cues, effectively
band-attenuating filters, e.g. depending on the frequency
resolution of the hearing impaired user. In an embodiment, the
notches (or peaks) extend over more than one frequency band in
width, e.g. more than 4 or 8 bands. In an embodiment, the notches
(or peaks) are in the range from 100 Hz to 3 kHz in width, e.g.
between 500 Hz and 2 kHz.
[0084] FIG. 8 shows embodiments of a listening device comprising an
ear-part adapted for being worn at an ear of a user, FIG. 8a
comprising a BTE-part comprising two microphones, FIG. 8b
comprising a BTE-part comprising two microphones and a separate,
auxiliary device comprising at least a third microphone. In FIGS.
8a and 8b, the face of a user 80 wearing the ear-part 81 of a
listening device, e.g. a hearing instrument, in an operational
position (at or behind an outer ear (pinna) of the person) defines
a FRONT and REAR direction relative to a vertical plane 84 through
the ears of the user (when sitting or standing upright).
[0085] In the embodiment of FIG. 8a, the listening device comprises
a directional microphone system comprising two microphones 811, 812
located on the ear part 81 of the device. The two microphones 811,
812 are located on the ear-part to pick up sound fields 82, 83 from
the environment. In the scene of FIG. 8a, sound fields 82 and 83
originating from, respectively, REAR and FRONT halves of the
environment relative to the user 80 (as defined by plane 84) are
present.
[0086] FIG. 8b shows an embodiment of a listening device according
to the invention comprising the listening device of FIG. 8a. The
microphone system of the listening device in FIG. 8b further
comprises a microphone 911 located on a physically separate device
(here an audio gateway device 91) adapted for communicating with
the listening device, e.g. via an inductive link 913, e.g. via a
neck-loop antenna 912. In the scene of FIG. 8b, sound fields 82, 83
and 85 originating from, respectively, REAR (82) and FRONT (83, 85)
halves of the environment relative to the user 80 (as defined by
plane 84) are present. The use of a microphone located at another,
separate, device has the advantage of providing a different
`picture` of the sound field surrounding the user.
[0087] The invention is defined by the features of the independent
claim(s). Preferred embodiments are defined in the dependent
claims. Any reference numerals in the claims are intended to be
non-limiting for their scope.
[0088] Some preferred embodiments have been shown in the foregoing,
but it should be stressed that the invention is not limited to
these, but may be embodied in other ways within the subject-matter
defined in the following claims. For example, in the described
embodiments, reference is generally made to two directions, FRONT
and REAR. Other directions than FRONT and REAR relative to a user
could be used depending on the application in question. Further,
more than two directions may be used without deviating from the
general concepts of the present invention.
REFERENCES
[0089] US 2007/0061026 A1 (Wang) 15 Mar. 2007 [0090] EP 1 443 798
A2 (PHONAK) 4 Aug. 2004 [0091] US 2007/230729 A1 (OTICON) 4 Oct.
2007 [0092] US 2009/0074197 A1 (OTICON) 19 Mar. 2009 [0093] WO
99/09786 A1 (PHONAK) 25 Feb. 1999 [0094] U.S. Pat. No. 5,473,701
(AT&T) 5 Dec. 1995 [0095] EP 1 579 728 B1 (OTICON) 8 Jul. 2004
[0096] [Middlebrooks, 1999] Middlebrooks, J. C., Individual
differences in external-ear transfer functions reduced by scaling
in frequency", J. Acoust. Soc. Am., Vol. 106 (3), pp. 1480-1492,
1999. [0097] [Blauert et al., 1997] Jens Blauert, John S. Allen,
Spatial hearing: the psychophysics of human sound localization,
Edition: 2, revised, 494 pages, Published by MIT Press, 1997, ISBN
0262024136, 9780262024136. [0098] [Wang, 2005] Wang, D. On ideal
binary mask as the computational goal of auditory scene analysis,
Divenyi P (ed): Speech Sepearation by Humans and Machines, pp
181-197 (Kluwer, Norwell, Mass. 2005).
* * * * *